Rehearse Tomorrow's Hard Conversation Tonight

Daily news for curious minds.

Be the smartest person in the room. 1440 navigates 100+ sources to deliver a comprehensive, unbiased news roundup — politics, business, culture, and more — in a quick, 5-minute read. Completely free, completely factual.

The conversation you need to have happens in your head every night.

Not the words — the dread. The replay. The catastrophic version where they say the worst possible thing and you say the wrong thing back.

What if you could actually have it tonight? Not with the actual person — with their pattern. Their pushback. Their tone. Their pause. You can. AI just got good enough at impersonating people from limited context that "rehearsing the conversation" stopped being a metaphor.

TLDR

The prompt below interviews you about who you need to talk to, builds a working persona based on what you know, then plays them in real-time. You can run the conversation cold. Watch where it derails. Try a second take. Walk into tomorrow's actual meeting having already had it three times. Voice mode makes it almost too real.

The Real Problem With "Preparing"

Most "preparation" for hard conversations isn't preparation. It's rumination.

You picture the conversation going badly. Then better. Then the worst version again. By the time the real meeting happens, you've lived 30 versions of it — none of which prepared you for the one that's about to actually happen.

The reason: your brain rehearses your lines, not theirs. You can't simulate someone else's reaction patterns just by thinking harder. You need a partner. Most of us don't have one — partly because rehearsing the conversation with a friend means leaking the situation, and partly because your friend isn't going to push back the way they will.

Why Roleplay Actually Works

When AI plays them — pushing back, going quiet, redirecting, getting defensive — you stop responding to the script you wrote and start responding to the conversation that's actually happening. Four things shift:

1 You hear yourself. The phrase you've been mumbling for two weeks suddenly sounds patronizing — or apologetic — or unclear.

2 You see the fork. When they push back at minute two, you find out what you actually do. Recover? Concede? Get defensive? Now you know.

3 You find the line. After three takes, the one sentence that lands becomes specific. The thing you couldn't quite articulate gets articulated.

4 You stop dreading. The conversation in your head is now in the rearview. The real one is just another rep.

The wow moment usually comes around minute three, when the AI says something the actual person would absolutely say — something you'd somehow forgotten was a possibility. That's the rep that saves you in the real conversation.

The Prompt

Run this in ChatGPT, Claude, or Gemini. Voice mode is even better — it forces you to speak the words instead of typing them, which is the whole point.

You are an expert in hard conversations and human dynamics. We are going to rehearse a difficult conversation I need to have.

Before any roleplay starts, INTERVIEW me with these questions ONE AT A TIME, waiting for my answer before asking the next:

1. What's the conversation you need to have? Describe it in one or two sentences.

2. Who is the other person? Their role, their relationship to me, and how long we've worked together or known each other.

3. What do you know about how they communicate? How do they react under pressure? What do they care about? Where do they tend to push back?

4. What outcome do I want from this conversation? What does the worst version look like?

5. Is this one isolated conversation, or part of a longer relationship I have to maintain afterward?

After my answers:
- Briefly summarize the persona you'll play in 3-4 sentences, so I can correct anything that's off.
- Ask if I want to start the conversation cold (you open) or have me open.
- Then ROLEPLAY as that person, staying fully in character.

While in character: push back like they would. Don't soften it. Don't make it easy. Use their likely tone, their likely deflections, their likely emotional reactions. If I say something that would actually land for them, react like the real person would. If I say something that would miss, miss like they would.

When I say STOP or DEBRIEF, step out of role and give me:
- 3 things I did well (specific phrases or moves)
- 2 things to adjust before the real conversation
- 1 sentence I should rehearse and have ready as a reset line if it goes sideways

Then ask if I want to run it again from the top, jump back to a specific moment, or be done.

Below: what came out of running this prompt across four very different conversations.

Your Boss Will Think You’re an Ecom Genius

Optimizing for growth? Go-to-Millions is Ari Murray’s ecommerce newsletter packed with proven tactics, creative that converts, and real operator insights—from product strategy to paid media. No mushy strategy. Just what’s working. Subscribe free for weekly ideas that drive revenue.

The AI That Scores Your Ads Before You Spend a Dollar

1 billion ad creatives generated. Used by Snap, Philips, Reckitt, and 4M+ marketers. Generate banners, product shoots, and video ads in seconds — and get a conversion score on every one before you launch.

Generate Your First Ad Free →

Mother's Day promo: 6 months free on yearly plans. Ends in days.

Weekly.ai BotBuilders

Where A.I. Meets ROI

PROMPT PROOF

Same Prompt. Four Conversations. Four Surprises.

What the AI surfaced that the user hadn't seen coming.

CONVERSATION	WHAT THE USER PLANNED TO SAY	WHAT THE AI ROLEPLAY SURFACED	THE FIX
The Raise Ask Senior PM → VP	"I think I've earned a raise. Here's what I've delivered this year."	AI-as-VP responded: "What number are you actually asking for?" — user had no number ready. Three takes later: still no number ready.	Walk in with a specific number and a justification. Without one, "I think I've earned a raise" becomes "let me think about it" and the conversation is over.
The Performance Talk Director → Underperforming PM	"Your work hasn't been at the level we need. We have to talk about it."	AI-as-PM went defensive at the word "level" — read it as a personal attack, started listing context. The framing triggered the spiral.	Lead with specifics, not labels. "The Q2 launch slipped twice on dates I owned" is processable. "Your work isn't at the level" feels like an attack on identity.
The Vendor Breakup Founder → 4-year contractor	"We're moving in a different direction. I appreciate everything you've done."	AI-as-contractor pushed for specifics ("Different how?") and offered concessions. Founder kept softening, opened a negotiation that didn't need to exist.	Make it a decision, not a discussion. "This is a decision, not a negotiation. I want to handle the transition well — here's what I'm proposing." Closes in 90 seconds.
The Personal Boundary Adult child → Parent	"We need some space. Please stop showing up unannounced. Here's why…" (and then a list)	AI-as-parent latched onto the "why" list and dismantled each item. The boundary got buried under 12 minutes of justification debate.	Stop explaining at item one. "Please call before coming over. That's the ask." The over-explainer always loses the boundary they were trying to set.

Same prompt. YOUR conversation. Try it. The persona AI builds for your specific person — the one you actually need to talk to — will surface things you've forgotten were possible. That's the rep that saves you tomorrow.

The Bottom Line

You've already had the conversation tonight. In bed. Eight times. The cost was sleep, not preparation.

Have it once tomorrow morning, with AI, on purpose. Then have it for real.

About This Newsletter

AI Super Simplified is where busy professionals learn to use artificial intelligence without the noise, hype, or tech-speak. Each issue unpacks one powerful idea and turns it into something you can put to work right away.

From smarter marketing to faster workflows, we show real ways to save hours, boost results, and make AI a genuine edge — not another buzzword.

Get every new issue at AISuperSimplified.com — free, fast, and focused on what actually moves the needle.